The installation Faces of AI reveals fascinating results of a large-scale image analysis by the LIT Robopsychology Lab: How is AI portrayed in public?
Actually, artificial intelligence has no body, no face. The current hype surrounding AI, however, demands visual embodiment. It demands images. The LIT Robopsychology Lab explores how AI is portrayed in public and what effect typical media images have on recipients. In the course of this research, 10,000 online stock images related to the keyword “Artificial Intelligence” were analyzed. These consisted of photos and illustrations, which are particularly used by the media. 118 students of the master’s program Artificial Intelligence at JKU Linz evaluated a random sample of 450 of the 10,000 images. Besides categorizing the motifs used (e.g. robot, avatar, brain, user, network), the students rated how beautiful they found an image, how creepy it felt to them, how human-like AI was depicted in it, and how realistic the respective representation of AI was from their expert perspective.
At the Ars Electronica Festival 2021, results of this study are presented to a larger audience for the first time in the form of the installation “Faces of AI”. Four polarizing filter eyes make the invisible visible. If visitors look through one of the eyes located in front of the installation’s screen walls, those images that were rated as particularly beautiful, creepy, humanlike or realistic by the students magically appear. The installation is also meant to inspire reflection: Which similarities are there between the images? What effect do the images have on observers? And what might alternative ways of depicting AI look like?
Credits
Project Lead: Martina Mara
Artistic Concept: Lisa Caligagan
Graphics: Laurenz Hintermayer
Technical Development: Lisa Caligagan
Technical Consulting: Christopher Lindinger
Data Retrieval & Analysis: Franz Berger, Martina Mara, Lisa Caligagan
